The proposed research is an in-depth, longitudinal study of a nonclinical sample investigating: (1) dynamics of daily alcohol consumption patterns over a 2-year period, and the biopsychosocial correlates of those dynamics; (2) natural history of criteria-based alcohol problems; and (3) associated dimensions related to identification of biochemical markers and specific risk factors for alcohol abuse and dependence. Goals: (1) Identify and quantify stable patterns of drinking, and (2) identify specific factors associated with perturbations in those patterns. Results from these studies could have a significant impact from both public health and public safety perspectives by: (1) specifying the conditions for the stability and change in drinking and symptom patterns (including alcohol dependence) over time, (2) providing a detailed prospective record of daily drinking, and (3) affording the opportunity to assess accurately the specificity and sensitivity of biochemical state markers of alcohol consumption: Also, these results will add substantially to scientific knowledge concerning: (1) subclinical symptomatology; (2) influences of life stressors on psychosocial functioning and alcohol consumption, and how these three classes of variables are interrelated; and (3) consistency and change in patterns of alcohol use. Method: Actively drinking male and female subjects recruited through newspaper advertisements will report consumption and other relevant information daily for 2 years using an automated Interactive Voice Response telephone system. Equal thirds of the subjects will be: (1) symptom free, (2) exhibit 1 to 2 DSM-IV alcohol abuse/dependence symptoms, and (3) manifest 3 to 5 symptoms. Daily data will be gathered on: quantity of beer, wine, liquor, and cigarettes; ratings of intoxication and while driving, general mood, stress, and physical health; and location of drinking (or reasons for not drinking). Subjects will be evaluated quarterly for development of or reduction in alcoholic symptomatology. Blood assays will be obtained every quarter for the 2-year period of daily reporting to determine the reliability and validity of two promising, recently-identified state markers of alcohol abuse (hemoglobin- acetaldehyde adducts and carbohydrate-deficient transferrin) relative to more traditional but nonspecific markers. Long-term objectives: Data obtained will be used to specify with greater reliability a number of highly relevant public health and public safety concerns. Since most Americans with alcohol problems are never treated, results of this study using subjects unselected for treatment-seeking should be especially meaningful for a broad spectrum of the drinking public. Data generated from this research should provide particularly useful insights into the evolution and fluctuations of symptomatology in a heavy drinking, nontreatment sample.